| With the development of multimedia technology, digital image is widely used in the fields of Internet, digital TV, visual telephone, network film, remote surveillance, medical photograph, household appliances, public security and so on. The contradiction is more conspicuous day by day between large number of data and the limited capacity of storage and band width. It is a hard task for the science and technology worker to do the research of image compression. Many significant achievements have been acquired as more and more theoretical results are obtained in the two decades. Based on the classics lossy and lossless image compression methods, various new algorithms are proposed to be applied to image compression, such as wavelet transform, fractal technology, artificial neural network(ANN). Higher compression ratio and better visual effect of image are achieved. Nevertheless, it is arduous for the new algorithms are applied to the real time fields on account of their complicated computation procedure. It is urgent to accelerate the computation speed of image compression.The standard organizations of JPEG(Joint Photographic Experts Group), MPEG(Moving Picture Experts Group) and VCEG(Video Coding Experts Group) have promulgated several international standards in accordance with various applications, in which the new efficient research achievements are adopted. The standards have recommended data styles and algorithms, meanwhile the endeavor direction of research workers is showed clearly.The general way of image compression is described in this thesis. We have analyzed the recent international standards, especially the image compression methods based on wavelet transform. We have discussed three typical algorithms. Several new video compression themes based on wavelet transform are proposed and the simulation effects were presented in the thesis. Even more we have probed into the subjects of fractal theory and ANN technology. At last we have introduced the design approach of the real time video compression board based on the ADV611 chip.The general way of image compression is described in the preface of chapter 1.We discussed the image compression international standards of MPEG-4 H.264 and JPEG2000 in chapter 2.In chapter 3, we introduced the theory of wavelet analysis and its recent progress, include multiresolution analysis, Mallat algorithms, biorthogonal wavelet, waveletpacket, multi-wavelet analysis. The lifting scheme was discussed too. We analyzed "Embedded Zerotree Wavelet Encoding", "A New, Fast, and Efficient Image Codec Based on Set Partitioning in Hierarchical Trees" and "Embedded Block Coding with Optimized Truncation of the embedded bit streams " algorithms in chapter 4.The vector quantization of wavelet coefficients is analyzed in chapter 5. The LBG algorithm is compared with SOFM vector classified searching method. A new vector classified searching method is proposed in this chapter that can accelerate searching speed with less influence to the visual quality. We presented a new video compression method based on MC (motion compensation) mainly the fast searching method of the motion vectors is described. In the end of this chapter we presented a new video compression algorithm of vector coding based on wavelet transform where the MC is combined with vector searching. Simulation shows high compression rate is achieved with less time-consumption.In chapter 6, We probed into the principle of fractal theory. The fractal image compression algorithm of Jacqain AE is analyzed and a new fractal image compression algorithm based on wavelet transform is presented.In chapter 7, we introduced the design approach of the real time video compression board based on the ADV611 chip. The static image compression technique of AD V611 being conjoined with the interrelationship between neighboring frames of video pictures, the DSP+FPGA hardware structure being adopted, the real time compression of video comes into reality.We made a summary of this thesis in the last chapter, and then the endeavor d...
|
PDF Full Text Request